Increase lscratch allocation for RemoveBL

[kelly-sovacool]

@lobanov

I had some failures with "RemoveBL" step, presumably running out of lscratch space (third row). Can we increase it in the next iteration?

[kopardev]

This will be addressed as part of issue #127 (implement attempt-based resource scaling for RAM, walltime, and lscratch).

Rather than a static bump to cluster.json, the proper fix is to make lscratch scale with the Snakemake attempt number — e.g. 256 GB on attempt 1, 512 GB on attempt 2 — so jobs that run out of scratch space automatically retry with more. See #127 for the full design.

_{⚡ Generated using AI ⚡}